What the Opening Weeks of 2026 Reveal

Key deals of January 1 - January 9, 2026

LMArena: The Layer That Decides Which Models Win

LMArena’s latest $150M round quietly reframes where power is accumulating in the AI stack. In just eight months, the company has tripled its valuation to $1.7B by turning real, crowd-sourced model evaluation into core infrastructure. LMArena (formerly Chatbot Arena) lets users anonymously compare systems like ChatGPT, Claude, and Gemini based on actual performance, not privately tuned benchmarks. Co-led by Felicis and UC Investments, with a16z, Kleiner Perkins, Lightspeed, and others participating, the bet isn’t on another model, but on the layer that determines which models actually matter.

“This new funding allows us to meaningfully scale our engineering, research, platform operations, and community initiatives to meet accelerating global demand,” LMArena said on X.

EpiBiologic: When Specificity Becomes the Edge

EpiBiologics’ $107M Series B lands as protein degradation moves from promise to precision. Rather than chasing broader EGFR inhibition the company selectively degrading cancerous EGFR while sparing healthy tissue and doing it on the cell surface, not inside the cell. That distinction matters. If EPI-326 delivers on its safety and specificity claims as it enters the clinic in 2026, it opens a cleaner path to combination therapy in cancers where resistance has become the norm.

Backing from GV, J&J, and Novartis’ venture arm signals confidence not just in a single asset, but in EpiBiologics’ EpiTAC platform as a new modality layer for oncology.

OpenAI & Convogo: Acqui-Hiring as Execution Strategy

OpenAI’s latest move is another reminder that its M&A playbook is less about products and more about people. By acqui-hiring the Convogo team in an all-stock deal and shutting down the product, OpenAI is pulling in founders who’ve spent the last two years translating raw model capability into usable, professional workflows. Convogo wasn’t acquired for its software, but for its lived experience at the edge where new model releases meet real human use cases. Per PitchBook data, It’s OpenAI’s ninth acquisition in a year.

The pattern is consistent: absorb teams that understand applied AI deeply, fold that insight into the core platform and use M&A as a shortcut to execution velocity in areas that matter most to its AI cloud ambitions.

ChatGPT Health: Turning AI Conversations Into Contextual Care Insights

OpenAI is carving out a dedicated lane for health in ChatGPT with the launch of ChatGPT Health. The product creates a separate space for users to discuss medical and wellness topics, keeping that context isolated from everyday chats while still leveraging prior interactions when relevant. It can pull in data from Apple Health, MyFitnessPal, and other wellness apps, but OpenAI emphasizes these conversations won’t train its models. The move is a response to persistent healthcare friction: cost, access, and continuity. ChatGPT Health isn’t a replacement for doctors, but it’s OpenAI’s latest experiment in turning AI into a more contextual, personalized assistant for complex human needs.

Across AI infrastructure, biotech and applied machine learning the edge isn’t always the flashiest product, it’s the layer, the specificity, the execution muscle and the people who translate capability into real-world impact. LMArena is defining which models matter, EpiBiologics is proving that precision compounds have an advantage, OpenAI is stacking teams that turn research into usable workflows and ChatGPT Health shows the payoff of contextual integration. 

In 2026 the winners won’t just move fast, they’ll build the foundation that makes speed sustainable.